Hot headed tank

[PaulAH]

There have been several posts about CH header tanks misbehaving, and useful replies, but could someone answer a basic question: How hot, if at all, should a header tank get in an open-vented system?

This morning I went into the loft to investigate water dribbling down an internal wall. It's coming from the header tank, which was hand-hot and lying in a puddle of water. Further investigation is needed as I don't know if the system is blowing back, or if there's HW/steam spitting out of the vent onto the floor, or if the tank has simply developed an old-fashioned leak (the CH had gone off by then and I'm now in the office).

But should the tank be warm/hot in the first place? That might provide a clue. I guess convection effects will push some heat into the tank and this one only holds about 3 gallons.

Paul

 

[PaulAH]

Glad everyone found that so exciting

The leak turns out to be just that, a leak (loose washer). But I've noticed another odd symptom - about 30 seconds after firing up the boiler from cold, the vent pipe squirts about a cupful of water into the header tank before settling down. Is this normal?

 

[PEDANTICVINDICTIVEMAN]

Vent pipe shouldn't really ejaculate, but some seem to do this, vent pipe is probably not high enough before bending and terminating over tank, orrrrrrrrrrrrr, could be incorrect coldfeed/vent arrangment or cold feed blocked.

 

[ChrisR]

If the cold feed goes downwards from the heating circuit then back up, to make an anti convection loop, the header tank stays pretty cold.

I'm really not sure why some systems pulse a spurt of water out of the vent, but having a bit of air in the system somewhere, which would squash and expand, would allow an extra burst of water speed as the pump starts and stops.

 

[PaulAH]

An engineer friend writes:

"When the pump starts it has to get a large body of water moving round the system. There's a pressure difference due to inertia between the near and far ends, and the line of least resistance is the vent pipe. So you'd expect the water level to rise there, perhaps to the point where some syphons out."

The brow of the vent pipe is about 10ins above water level. Does this all sound normal?

 

[stevie-c]

If the cold feed goes downwards from the heating circuit then back up, to make an anti convection loop, the header tank stays pretty cold.

I'm really not sure why some systems pulse a spurt of water out of the vent, but having a bit of air in the system somewhere, which would squash and expand, would allow an extra burst of water speed as the pump starts and stops.

Is a non return valve on the cold feed an acceptable method of stopping convection.

 

[PEDANTICVINDICTIVEMAN]

18" high vent above water level is the norm for some systems, look in your boiler makers installation instructions.

 

[oilman]

Non return valves in the feed and EXPANSION pipe is not acceptable. Water has to flow both ways.

 

[PaulAH]

18 inches does sound more sensible given the size of the system (24 rads). I'll try raising the vent. The manufacturer's boiler blurb gives no information on this.

 

[brumylad]

I thought there was a formula to predict the height but I had a look to check before this post and as far as I can see this only applies in the case of a hot water cylinder vent.

 

[wkb21]

But should the tank be warm/hot in the first place? That might provide a clue. I guess convection effects will push some heat into the tank and this one only holds about 3 gallons. Paul

NO it should be cold. I have this problem, and it's not due to expansion pipe overflow. The height of my F&E tank above the point where the cold feed pipe joins the delivery line from boiler to the heat load is only about 1m. The tee off the delivery line has the feed from the boiler on the leg of the tee, with feed line coming in on the upper side of the T (imagine a letter T on its side with boiler coming in on R, delivery to load downwards and cold feed at top). I queried the wisdom of this when the installer (fully certificated but clearly not wise) put the boiler in. He was adamant that it was standard procedure!

The result is that delivery from the boiler slams into the leg of the tee and creates eddies up the cold feed line. Natural convection (even in a 15mm pipe) does the rest and drives a circulating loop up and down the feed pipe into the F&E tank. As a result my F&E tank sits almost at boiler delivery temp.

I discovered this 2 days ago. Rather than go back to a less than competent plumber I plan to install an anti-convection bend INSIDE the F&E tank using rubber tubing on a support (at the cold feed outlet hole) with the top of the U just below the top of the tank and the tube open end at the level of of the cold feed outlet hole. This should be adequate to break the convection loop and allow free expansion and feed into the cold feed pipe via the rubber tube. Free passage of water will be unimpeded for pressure relief, feed and expansion purposes.

I have the impression that there are lots of competent plumbers, but not many that understand well enough to anticipate these problems, or solve them.

Any comments? I'll report progress.

 

[45yearsagasman]

wkb21 You obviously have no idea on system design and what you propose to do to your expansion tank is cobblers.

 

[wkb21]

wkb21 You obviously have no idea on system design and what you propose to do to your expansion tank is cobblers.

Thanks for your expert appraisal and diplomacy. Please explain to us all why my diagnosis and solution is cobblers.

 

[DP]

WKB21, I think you must have been hibernating all this time. The post was written 14 months before you joined the forum.

Would not be surprised if the poster has moved since.

 

[wkb21]

Any comments? I'll report progress.

After a couple of days of testing and fine tuning I'm pleased to report that it works, and I now have a cold header tank. The anti-siphon loop also needs to be high enough to overcome the pressure surge when the pump starts up. This will depend on pump speed and height of expansion tank above the pump. Look at the pump characteristics to check what the max head is at zero flow at the speed you're set at.